Skip to content

← Psychology

Conformity

Chapter 22 · Other people

In this chapter 4 sections
  1. Two reasons to go along
  2. Asch’s lines
  3. Seventy years on, and across the world
  4. Changing your mind to match what you did

You arrive in a town you’ve never visited, hungry, and there are two momo shops side by side. One is packed, with people waiting on the steps. The other is empty. Most people walk straight to the crowded one, and they’re usually right to. Fifty strangers made a choice before you, and their feet are telling you something.

Now a different scene. The teacher asks the class a question and gives three answers to choose from. You’re fairly sure it’s the first. Then hand after hand goes up for the second, and when your turn comes, your hand goes up for the second too.

Both times you went along with other people. Psychologists call that conformity: changing what you do or say so that it matches a group. But the two scenes are not the same, and the difference between them is most of what this chapter is about.

Two reasons to go along

The first reason is that other people know things. When you’re unsure, what everyone else is doing is evidence, often good evidence. The packed momo shop is a kind of report, from fifty people who had no reason to mislead you. Psychologists call this informational influence: you go along because you think the group is probably right. It isn’t a weakness. People learn most of what they know this way, and a mind that ignored everyone else would have to find out from scratch which mushrooms are poisonous. Mathematical models of how people learn from one another have found that copying the most common choice helps them home in on what works in the place where they live. When volunteers were tested on a computer task, they leaned toward the majority mainly when they were unsure of themselves and when many people agreed, which is just when copying makes sense.

The second reason is that you want to belong. Standing out alone is uncomfortable. You might be laughed at, or thought odd, or left out. So you go along even though you think the group is wrong, and you keep your own view to yourself. This is normative influence, from the word norm, the unwritten rule of what people around you do. It is about being accepted, not about being right.

So the momo shop is mostly the first kind, and the raised hands are probably a mix of both. In real life the two are tangled together, and even in the laboratory they are hard to pull apart. One way to tell them apart is to ask what you believe afterward, in private. Did you change your mind, or only your words? Going along in public while still disagreeing inside is sometimes called compliance, to set it apart from really being persuaded.

Asch’s lines

The experiment everyone remembers took away most of the first reason, so that mostly the second was left.

In the early 1950s the American psychologist Solomon Asch, who taught at Swarthmore College, invited college students to what he called a test of vision. A group of seven to nine young men sat together in a classroom. The experimenter held up two white cards. One had a single line on it. The other had three lines, and the men had to say, one by one and out loud, which of the three was the same length as the first.

the standard line123which one matches?
Figure 22.1 The task, drawn from the first pair of cards in Asch’s own list: a 10-inch line on the left, and lines of 8¾, 10 and 8 inches on the right. Line 2 matches.

It was easy. On the rounds that mattered, the wrong line the group would pick was off by about 2 to 4.5 centimeters, and people judging alone got it wrong less than one time in a hundred. The first two rounds passed quietly, everyone agreeing. Then, on the third round, everyone who answered before one particular man gave the same wrong answer.

That was the trick. All of them except one had been told beforehand to give wrong answers on 12 of the 18 rounds. The one who hadn’t, seated near the end, was the real subject. He had the evidence of his own eyes on one side and a calm, united group on the other, and he had to answer out loud in front of them.

Asch tested 123 young men from three colleges this way. On the rounds where the group was wrong, they went along with it in 36.8% of their answers. People differed hugely. About a quarter never once agreed with the wrong answer. At the other end, six men went along every single time. And each man tended to keep to his path: those who held out early mostly held out to the end, and those who gave in found it hard to stop.

Asch talked to every one of them afterward, and their answers are the most human part of the work. Some of those who held firm were simply confident. Others came to believe the group must be right, yet kept giving their own answer because, as they put it, they had to call it as they saw it. Among those who gave in, some had quickly decided “I am wrong, they are right”. Others said they went along “not to spoil your results”. Many suspected the others were following the first speaker like sheep, or were fooled by some trick of the light, and still went along at the moment of answering. The ones Asch found most troubling had decided that disagreeing meant something was wrong with them, a flaw to hide at any cost. And every one of those who gave in thought he had given in less often than he really had.

This study is usually told as proof that people will deny their own eyes to fit in. That’s half of it. The same numbers say that almost two answers in three were right, under pressure, and a quarter of the men never gave in at all. Asch himself cared as much about the ones who stood firm. He was worried that “reasonably intelligent and well-meaning young people are willing to call white black”, but he ended his report by saying that people’s power to stand alone should not be underestimated, and that nearly everyone in the experiment agreed afterward that independence was better than going along.

A reader of the article wrote in to argue that Asch had it backward. A good scientist, he said, gives up his own impression when the evidence goes against it, and the agreement of everyone else in the room was the best evidence a person had. So going along was sensible, and digging in was stubbornness.

It’s a fair point, and it is really the first reason to go along, the momo shop, put as a rule. Asch’s answer was just as fair. The man in the experiment wasn’t asked what the others saw. He was asked what he saw, like a member of a jury, who must weigh what the others think but can’t hand his own responsibility over to them. And history, Asch wrote, shows plainly “that error and distortion can have quite a following”. Agreement isn’t proof.

So both are true, and the skill is in telling the situations apart. When you’re unsure and many people who know more than you agree, going along is usually wise. When you can see clearly for yourself, the crowd’s agreement adds nothing, and saying what you saw may be the most useful thing you can do, for the group too.

Asch didn’t stop at one experiment. He changed one thing at a time to see what mattered.

First, how many people were against you. With only one person giving the wrong answer, almost nobody went along. Two made a real difference, three made more, and beyond that adding people changed little. Fifteen were no more persuasive than four.

10%20%30%40%0alone: under 1%123467915people giving the wrong answer
Figure 22.2 Wrong answers given by Asch’s subjects as the number of people giving the wrong answer grew. Judging alone, people were wrong less than 1% of the time. From his 1955 report.

Second, and far more powerful, was having one ally. When just one other person in the room gave the right answer, whether another real subject or a planted helper, the subjects went wrong only a quarter as often. The group’s power came less from its size than from its being united. Even a person who broke from the group in a different direction, choosing a line still further off than the group’s, freed people: with such an extreme dissenter in the room, their errors fell to 9%.

Third, the ally had to stay. When a supporting partner switched sides halfway through and joined the majority, the subjects gave in about as often as if they’d been alone from the start. When the partner simply left the room, saying he had an appointment with the dean, the support he’d given lasted a little after he was gone.

Fourth, how wrong the group was. Asch hoped to find a gap so large that nobody would give in. He didn’t find one. Even when the lines differed by 7 inches, about 18 centimeters, a few still went along.

Later studies, pooled together, added a few more. In the American ones, people conformed more when the task was harder or less clear, which fits the first reason: the less you can trust your own eyes, the more other people’s answers are worth. They conformed less when the majority was a group they didn’t belong to. And one result went against what the researchers had expected: whether the group would learn your answer made no clear difference.

Seventy years on, and across the world

Asch’s studies were done on American students in the 1950s, and some writers have suggested that the political fears of those years in America made people go along more than usual. Would it still work today, somewhere else?

In the spring of 2021, two sociologists at the University of Bern in Switzerland ran it again with 210 people recruited on their campus, men and women. Five planted helpers, one real subject, ten rounds of lines, six of them with the group giving the wrong answer. On those six, people went along 33% of the time, close to Asch’s figure.

Then they added something Asch hadn’t. Half the people were paid one Swiss franc for every right answer. The money helped, but not much: the error rate fell to 25%. Even when being right paid, a quarter of the answers still followed the group.

They also tried opinions. After the lines, each group answered five yes-or-no questions about Swiss politics, such as whether trade unions should have more power, with the helpers all answering one way. When the group said yes to more power for the unions, 72% of the real subjects said yes. When the group said no, only 29% did. Across the five questions, what the group said shifted people’s answers by 38 percentage points on average.

And the personality tests they gave afterward, of intelligence, self-esteem and the need to be approved of, didn’t predict who would go along. Only one trait, openness to new experience, went with going along less, and the team asked for that to be tested again before anyone trusts it.

If the second reason to go along is wanting to belong, you might expect it to be stronger where belonging is prized more.

In 1996 two psychologists at the University of Sussex in England gathered 133 studies that had used Asch’s lines or something close to them, from 17 countries. In the American studies, conformity had gone down steadily since the 1950s. Across countries, the more a country’s people put the group before the individual, by three separate surveys of values, the more people in its studies tended to go along. These are sometimes called collectivist cultures, against individualist ones such as the United States.

That is a tendency, not a rule, and it needs care. The countries were scored by surveys of other people, not of the students in the experiments. Ninety-seven of the 133 studies were American. None was from Nepal, or even from India; the closest to a South Asian sample was in Fiji, where one study found that people of Indian descent went along more than native Fijians. The researchers themselves added a thought that matters a good deal for a reader here. In a culture that values harmony, agreeing in public while privately disagreeing may be seen as tact, as good manners, rather than weakness. Called by that name, they wrote, what the experiment measures might be better described as social sensitivity, and standing out as tactlessness. The word “conformity” carries a Western judgment with it.

So the same act can mean different things. Whether giving way to the group in Nepal, before elders or in a classroom, is more often tact or more often fear is a question nobody seems to have tested here.

Changing your mind to match what you did

So far the group has pulled on what you say. A different idea, from the same years, asks what happens afterward. If you’ve said something you didn’t believe, does what you believe change to match?

In 1957 the American psychologist Leon Festinger put forward a theory he called cognitive dissonance. Holding two thoughts that don’t fit together, such as “I think this is boring” and “I just told someone it was fun”, is uncomfortable, and people tend to ease the discomfort by changing one of them. The thing you’ve done can’t be undone, so the belief is the one that moves. But only if you have no good excuse. If someone paid you well to say it, the money explains what you did, and the thoughts fit well enough.

With James Carlsmith he tested it in an experiment as odd as anything in psychology. Seventy-one men from an introductory psychology class at Stanford University each spent an hour doing deliberately dull tasks: putting 12 spools on a tray, emptying it and filling it again for half an hour, then turning 48 square pegs a quarter turn at a time for another half hour. Then the experimenter told each man, falsely, that the study was really about expectations. To most of them he added that the usual helper hadn’t come. Would they step in, and tell the next student waiting outside that the tasks had been fun?

Some were paid one dollar to do it, and some twenty dollars. A third group did the same tasks and were never asked. The waiting “student” was a young woman hired for the part, who said a friend had told her it was boring. Most men answered with something like “Oh, no, it’s really very interesting”. Afterward each man was interviewed, apparently for a quite separate survey of the department’s experiments, and asked how enjoyable the tasks had really been.

Never asked−0.45Paid $1 to say it was fun+1.35Paid $20 to say it was fun−0.05−2−10+1+2← dullenjoyable →
Figure 22.3 How enjoyable the men later said the dull tasks were, from −5 (extremely dull) to +5 (extremely enjoyable), 20 men in each group. Festinger and Carlsmith, 1959.

The men paid twenty dollars rated the tasks about as dull as those never asked. The men paid one dollar rated them clearly more enjoyable. A single dollar was too little to explain the lie, the theory said, so they had come round, a little, to believing it. Eleven of the 71 weren’t counted: five who suspected the trick, three who refused the money, two who told the woman they had been hired, and one who asked for her phone number so that he could call and explain.

That result became one of the best-known in psychology, and getting people to say what they don’t believe became one of the most popular ways of studying the theory. In one version, people write an essay arguing for a view they disagree with, such as higher fees at their own university, and some are told it is entirely their choice to do so while others are simply told to. The theory says those who felt they chose should change their view more, since they have no excuse.

In a test published in 2024, 39 laboratories in 19 countries, working in 14 languages, ran that essay version with 4,898 people. The result did not support the theory’s central prediction. Feeling free to choose made no clear difference to people’s views afterward. Those who wrote the essay did end up with different views from those who wrote about something neutral, but the authors concluded that their results call into question whether this kind of experiment gives solid evidence for cognitive dissonance.

So one of the theory’s most famous kinds of evidence failed a large, careful test. That is not the same as showing that people never adjust their beliefs to fit what they’ve done. The 2024 study didn’t repeat Festinger and Carlsmith’s dollars, and dissonance has been tested in other ways it didn’t try. But the experiment that was supposed to show it most plainly didn’t, and whether the discomfort Festinger described changes minds is open again.

Going along with other people isn’t one thing. Sometimes it is learning, the sensible use of what others know, and sometimes it is the wish to belong, which can carry you away from what you can plainly see. Asch’s lines showed how strong the second pull can be when a group stands united, and how much weaker it becomes when even one person breaks ranks. That has held up across seventy years and many countries, though its size changes with the times and the place, and even what to call it depends on what a culture values. Changing your beliefs to fit what you’ve said is a shakier story. The next step is a harder case still: not a group of equals, but one person with authority, telling you what to do.

Found a mistake? Please tell us.